Hello all, wondering if you could help me out with understanding the basics of fluorescent tubes. Just let me know what number question you are answering so I don’t get confused lol.

#1: Instant start

Do these ballasts use a high voltage pulse to ignite the fluorescent tube or do they just have a very high open-circuit-voltage that can strike the lamp without preheating?

#2: Single-pin tubes

Instant start ballasts can obviously either use the conventional fluorescent tubes with 2 pins on each end or use the special ones with only 1 pin on each end. With the single-pin tubes, is there still a filament inside that is just connected together or is it more like a neon sign electrode? I would assume the lifespan would be very long with solid neon-like electrodes.

#3: Dark ends and lifespan

What actually causes the very dark ends and low lifespan in an instant start configuration? Is it just the mere act of igniting with cold electrodes or is it a high voltage pulse that blasts away the material?

#4: Preheat vs Rapid tubes

I hear that there are different fluorescent tubes for preheat and rapid-start ballasts. What would really happen if you mixed them up? Is there any backwards compatibility?

#5: 2 pin PLs

Some PL lamps have 2 pins, while others have the sensical 4. Are the 2 pin PLs an instant start type? I don’t have a lot of PL experience.

#6: Stupid wattages

This is absolutely despicable in my eyes lol. There are F14T12s and F15T12s (stupidly close). Also F13T5 and F14T5. Why? Just to be annnoying? Also why are circline tubes always special wattages? Why not just have a bunch of pre-set wattage increments that everyone follows to simplify ballasting. Like have a 20w 1” diameter circline, a 20w 1” diameter u-bent, and a 20w 1” diameter linear that can all use the same ballasts. Obviously there is no turning back now, but it would’ve been smart to start that way.

#7: Aircraft T5s

Well of course, just to add to the ambiguity, aircraft tubes have to have different ends than other lamps. There seem to be normal sizes available (F4t5, F14T5 etc), but the ends are different. Is that the only difference? Could you just hook it up to your everyday ballast that matches the lamp wattage and size and have it work (assuming you could find the sockets)?

#8: Fulham workhorse…

How the actual heck do these work? How do they know the specifications of the lamp that they are powering?

Sorry for asking so many questions, thanks for your help! 

1 till 3: Instant start use higher OCV of the ballast. It is not only to initiate the plasma, it also needs to provide sufficient power in the cold cathode (so high cathode voltage drop) mode so the cathodes do warm up quickly enough, because it is the operation at cold cathodes which sputters off the electrode coating so wears of the cathodes.

4. Preheat circuit needs the filaments resistance to be high enough so just mildly elevated ballast current (e.g. about 600mA on an otherwise 430mA F40T12) is sufficient to warm the electrodes to the emission temperature in a reasonable time (normal starters give barely 200ms per cycle, so usually need about 2..3 cycles to reach emission for start). This higher resistance then forms extra losses during normal operation (so reduce the final lamp efficacy) On the other hand RS ballasts feed the filaments from low impedance windings, so allow easily to feed the filaments with way higher current than the normal arc current, so design the filaments with lower resistance so lower losses during normal burn.

5. 2-pin PL's are intended to operate at a basic series choke ballast, they integrate the lamp and starter into one unit. Normally when the tube is dying, it uses to significantly wear the starter, the worn off starter then prematurely wears off the new tube, so you end up with very short lifetime. So both need to be replaced together (except starters explicitly rated to last over multiple tubes). So by having both together it is ensured a new fresh starter will come with each new tube, without any need for the user to care that much.
Then when electronic ballasts came, the circuit was not using the starter anymore, so the 4-pin tubes were introduced.

6. The wattage is quite a result of the other design objectives, which with fluorescents start with "where it needs to fit". The main motivator for a lamp type is then the shape (linear, circular, U,...), length (it has to fit given ceiling tiles or furniture width formats) and surface brightness (sometimes low brightness is needed). Then it has to use as much build components as possible, mainly those which wear off or use to fail frequently (mainly sockets, starters), where the sockets dictate the diameter. The length and brightness gives the power, the length and diameter the arc voltage, so with the given power the current.
So you may have two tubes with very similar power, but when one needs to operate at low brightness and needs to be long, while the other needs to be shorter and higher intensity, you end up with way different arc voltages and currents. It is not possible to unify the electrical parameters of such two lamps, the physics just does not allow it.
And because the ballast uses to be quite long living component, it is not considered as that big of an issue if there need to be so many of them.
In the last years of fluorescents (the high efficency T5HE and T5HO era) there is visible a tendency to unify at least the arc current, e.g. all T5HE lamps are rated at 165mA, so it becomes feasible to support quite wide range of lamp types (even when not all) by a single ballast type, even in the areas with restrictions like EU or so (where ballast must drive the lamp so it emits its rated lumen output) so where ballast underdriving lamps (the way how WorkHorse "achieves compatibility" with so many lamp types; what matters is not the power, but the lamp light output) are illegal.

7. Aircraft use needs the lamp to be high shock and vibration resistant, so require rather beefy and solid sockets. On the other hand the "civilian" use requires the connectors to cheap to make, be simple to assemble and provide quite strong protection against accidental touching of the live conductors (aircraft lighting, as anything else on aircrafts is serviced by specially trained professionals only, while the domestic lighting needs to be "redneck-Joe proof").

8. The WorkHorse just does not operate all the "compatible" lamps at the correct rated power, it uses to be even 50% below rating. The point with electronic ballasts is, it is not that difficult to make the characteristic so, it wears the cathodes in cold state for really minimum time (either proper preheat, or maintaining really high current in cold cathode mode so the electrodes warm up quickly), so it then gets away with the power being way off the rating. For instance the 10..30W F10T8 till F30T8 all have the rated current in 0.22..0.38A range, so if the ballast delivers the 0.22A even at high arc voltage, all lamps work there pretty long life, although many at way lower than rated power. Because the way how the illumination is perceived, not many people notice that without side-by-side compare, so they get away with this approach.
However the fact the lamps are not driven according to their rating, such ballasts are illegal e.g. in Europe (the legislation does not allow any gear that leads to lower than rated lamp output or higher than rated lamp power).

Wow! You certainly know a lot (or maybe I don’t know enough)!

This was super helpful, thanks so much for your response!

So just to be clear:

• Instant-start systems keep the filament hot solely by the current flowing through the gas discharge
• When a lamp is discharging while it’s electrodes are cold, sputtering occurs
• Workhorse ballast just sort of run lots of lamps “good enough-ish”

So this begs the following question (I know, I am asking a lot):

#9: Solid electrodes

If the only reason instant-start lamps don’t last as long is because they start cold, so the filament eventually sputters away and breaks, why not just not have a filament? If you just had a sheet metal disc at each end of the tube, it would obviously sputter, but it would probably last decades before completely sputtering the material away. Like a CCFL in a laptop screen or a neon sign. Why not make them like this?

Thanks again for your quick response!

  The  slimelines  instant start do last long time.  Thet are use in  low switching on and  off location.   They were the workhorse of fluorescent in many stores where they are  started and left on  once a day or 24/7 use .   It the number of starts is what kills the cathodes on Instant start.

Medved's answers are always impeccable, but I can add a little more:

2. The single-pin slimline tubes indeed still have a hot filament cathode inside.  In the old days some manufacturers made them with neon-style cold cathodes and indeed the life is then longer, but efficacy is much worse because the voltage drop in front of a cold cathode is much higher than from a hot cathode.  Such lamps then have to be run on different control gear that offers a necessarily higher voltage.  Moreover the cold cathode electrodes only achieve longer life when run at quite low discharge currents.  It is not realistic to achieve the same high luminous flux from a cold-cathode fluorescent tube as the hot cathode types.

5. To answer your question about why there is an F15T12 and an apparently very similar F14T12, this builds on the arguments already given to solve a particular problem.  In this case the F15T12 was introduced first in April 1938, intended for use with a single lamp on a single choke on 110-120V mains supplies.  Very quickly there were desires to use fluorescent tubes in street cars which used to have a 64-volt supply.  The voltage of the F15T12 was too high, so at the end of 1938 the shorter F14T12 was developed with a lower discharge voltage.  Almost simultaneously, manufacturers of fixtures realised that fluorescent lighting was proportionally more expensive for the shorter tubes because the cost of the ballast is pretty much the same whatever its size - so the ballasts for the short tubes made up a higher percentage of the total fixture cost.  This could be solved by running two tubes in series on a single ballast.  Once again the voltage of two F15T12's in series would have been too high, but with the lower voltage F14T12 it was possible.  Incidentally the F14T12 was made shorter to reduce its voltage, but that was not quite enough.  It was therefore arranged to run it at a higher discharge current, which causes a further voltage drop of just enough magnitude that two tubes ran stably in series.

7. The aircraft lamps are also technically different in that they generally use different geometries of tungsten coils for the electrodes, which are better resistant to the persistent exposure to high frequency vibrations.  It is usually required to dim them quite deeply, when ordinary lamps may begin to flicker or snake and produce an undesirable appearance, so the gas filling is sometimes optimised to prevent such behaviour.  Usually they are also run either on the 400Hz AC supply circuit of the aircraft, or on high frequency transistor ballasts.  For optimal ignition behaviour and efficacy, the electrodes may be optimised for those conditions.  Usually the aircraft lamps can also be used in general lighting circuits, but not the other way around.  In addition to the special caps, many aircraft lamps are also made with the regular bi-pin caps combined with special vibration-resistant lampholders.

In some cases, there were 4 pin PL CFL lamps that used external starters and preheat fluorescent ballasts such as the 9 inch 18W PL-L for F20T12 ballasts and the 36W PL-L for F40T12 ballasts except for some lead-lag 2 lamp ballasts. Usually, they are used on programmed start electronic ballasts. Interestingly, Japan has been manufacturing and using a wide range of 4 pin CFL lamps that used preheat fluorescent ballasts and external starters. In many cases, they were often operated with a simple choke ballast on 100V mains, but some used autotransformer ballasts.

#9: Solid electrodes

If the only reason instant-start lamps don’t last as long is because they start cold, so the filament eventually sputters away and breaks, why not just not have a filament? If you just had a sheet metal disc at each end of the tube, it would obviously sputter, but it would probably last decades before completely sputtering the material away. Like a CCFL in a laptop screen or a neon sign. Why not make them like this?

Thanks again for your quick response!
[/quote]

The operating temperature of the electrodes is high, so it needs to be somewhat thermally insulated from the glass seal and the surface of the hot spot needs to be kept small, otherwise it woul loose too much heat, so consume too much power to just keep the electrodes warm.
Normally this is done by using just one rather small hot spot as the cathode (often called arc root). As the coating at that spot is wearing off, the cathode spot moves along the filament.

The fact the cathode is made as a filament or other shape, by itself does not make any difference on the wear. What wears off is not the metal filament itself, but the emission coating on its surface. If there is sufficient amount of it to form the root, the temperatures are low enough so the tungsten filament is not affected at all. Only after the emission material gets consumed, the arc root has no other choice but to sit on the tungsten itself, but then it needs to warm it up to way higher temperatures to keep the emission and it is this higher temperature what breaks the filament after the end of life. Attempting to make it "more solid structure" won't help, once the emission layer gets consumed, the electrode assembly would overheat anyway. Plus the fact the filament breaks so easily when overheated is rather welcome feature - on many ballasts (those which need to have the filaments conductive) it tends to act as an end of life fuse, so prevents the overheated electrode assembly from either shattering the tube. Of course this "protection mechanism" does not work with tubes connected by just a single wire at each end.

Medved's answers are always impeccable, but I can add a little more:

2. The single-pin slimline tubes indeed still have a hot filament cathode inside.  In the old days some manufacturers made them with neon-style cold cathodes and indeed the life is then longer, but efficacy is much worse because the voltage drop in front of a cold cathode is much higher than from a hot cathode.  Such lamps then have to be run on different control gear that offers a necessarily higher voltage.  Moreover the cold cathode electrodes only achieve longer life when run at quite low discharge currents.  It is not realistic to achieve the same high luminous flux from a cold-cathode fluorescent tube as the hot cathode types.

Ahh I see, so it is technically possible but not very practical or efficient.

In fact, cold cathode general lighting is (was) used from time to time. I personally spotted some in Moscow (high end clinic), Brussels (underground crossing) and Dusseldorf (Jewish center). High cathode voltage drop (~100V vs ~10V of hot cathode) is overcome by generally high driving voltage (~1kV) and long slim tubes, so efficiency penalty is not that bad, and even high-CRI phosphor coatings can be specified. But the main problem is that while regular tubes are still made in millions by multiple suppliers, cold cathode tubes are made to order mostly by hand, and while long-living, replacements availability is very questionable. 

More funny, some years ago, around 2000's the was an attempt of a German firm to make hot cathode sign tubes. What you got was about T5 HE 175 mA tubes, BUT in free form. Manufacturing process was adapted to what neon makers typically use with glass heating using a bombarder transformer, but these required a custom processing gas, not just 3mmHg air pre-fill and also a special power supply to process electrode filaments. It was cumbersome, and also hot cathode lamps are too fragile to rough life in the outside signs, so despite higher efficiency/brightness and nicer shorter electrode ends, this technology did not survive a competition with classic cold cathode sign tubes and and LEDs too and went bankrupt a few years ago. Pity I did not get last supplies at that time as making a custom hot cathode tube is definitely some fun!

 #7 | Me and Alex (@HIDLad001) ran a GTE Sylvania FA4T5/WW/RS/5004 on an Advance LQ-106F-TP (4-8w rapid start ballast) and it worked very well as there was cathode heating. But on a normal preheat circuit, they do nothing. Also for aircraft lamps 5000 series lamps have S14s bases and 5100 series lamps have normal G-5 bases-